We propose to test the hypothesis that the characteristic operational unit in the central nervous system is the module. This is conceived as an ensemble of neurons in a gliovascular stroma with a specified set of connections appropriate to the particular portion of the neuraxis. Each class of modules is believed to generate a logic peculiar to itself and a succession of such logical operations are conceived as the basis for information processing in the nervous system. Putative modules have been identified in spinal cord, thalamus, and cerebral cortex on the basis of structural (Golgi) techniques and we now propose to pursue this work in more detail with collaborative help from an electron microscopist as indicated. In addition to the usual methods of visual qualitative analysis of our Golgi material by drawing and photography, selected cell ensembles will be prepared for computer evaluation. The latter should prove useful in specifying quantitative aspects of connectivity patterns and in recognizing certain invariant aspects of domain geography where they exist. The functional aspects of neuropil modules will be explored with clusters of semi-microelectrode wires arranged to provide fairly high resolution extracellular recordings from groups of closely adjacent neurons. Experience with this technique in the chronically implanted animal indicates that the behavior of a number of units can thereby be studied simultaneously. We will use both visual and computer techniques in seeking evidence of regular firing relationships between elements of the unit ensemble during certain aspects of behavior. It is hoped that such relationships, when related to the position and nature of the responsible units, will provide insights into the kind of logic used by that module class in information processing.